Satellite communications (SatCom) are envisioned as a critical enabler of 6G networks, enabling seamless global coverage by integrating terrestrial infrastructures with multi-layered satellite constellations. Among these, the integration between geostationary (GEO) and low Earth orbit (LEO) satellite networks is particularly attractive, as they combine the broad coverage of GEO satellites with the low latency and high capacity of LEO systems. Within this context, we address the resource allocation problem for LEO satellite through a joint design of beam size and transmit power, while accounting for GEO interference constraints, residual Doppler frequency offsets, and frequency reuse strategies. The objective is to maximize the spectral efficiency of LEO system operating in multi-beam GEO-LEO networks. Motivated by the limitations of classical deep reinforcement learning (RL) in such dynamic orbital settings and the potential of quantum RL for accelerated convergence, we propose a hybrid solution that exploits quantum acceleration during offline training and subsequently exports the learned policy into a classical representational format for onboard LEO satellite deployment. A fully quantum deep deterministic policy gradient framework with variational quantum circuit-based actor and critic is developed, along with a neural network-based policy translator for classical inference. To the best of our knowledge, this is the first deployment-ready quantum RL framework in SatCom, offering efficient offline training, reduced retraining latency, and practical deployment compatibility with existing LEO satellite hardware.

Training federated learning (FL) at scale suffers from severe communication and heterogeneity constraints. These challenges are amplified in quantum federated learning (QFL), especially under non-IID data. We propose a model-driven QFL (mdQFL) framework that addresses communication overhead, scalability, and client drift through adaptive clustering and representative aggregation. The framework enables structured personalization and efficient update compression across rounds. mdQFL is the first QFL approach to jointly analyze training efficiency, personalization, and test generalization under heterogeneous conditions. Experiments across multiple datasets and quantum platforms show 50% communication reduction while maintaining or improving accuracy over standard QFL base-lines. We provide convergence guarantees and communication complexity bounds to establish scalability and robustness.

We propose AQM-LLM, a framework that distills Large Language Models for Active Queue Management in modern networks. Unlike conventional learning-based AQMs that require extensive feature engineering and struggle with dynamic conditions, our approach leverages the contextual reasoning capability of LLMs to enhance the Low Latency, Low Loss, and Scalable Throughput (L4S) architecture with minimal manual intervention. The L4S-LLM design introduces three key components: (i) a state encoder that transforms heterogeneous network telemetry into token embeddings, (ii) a specialized L4S-LLM head that produces congestion actions in a single inference step, and (iii) a data-driven Low-Rank Adaptation scheme that drastically reduces trainable parameters while preserving accuracy. Our open-source FreeBSD-14 implementation demonstrates improved queue delay stability and higher bandwidth utilization for both DCTCP and UDP Prague traffic. We emphasize that this work demonstrates architectural feasibility through controlled experiments; deployment on router-class hardware will require additional model optimization (e.g., compression, pruning, or quantization) and is left for future work.

This demo presents a modular experimental testbed and lightweight visualization tool for evaluating TCP congestion control performance in wireless networks. We compare Google’s latest Bottleneck Bandwidth and Round-trip time version 3 (BBRv3) algorithm with loss-based CUBIC under varying Active Queue Management (AQM) schemes, namely PFIFO, FQ-CoDel, and CAKE, on a Wi-Fi link using a commercial MikroTik router. Our real-time dashboard visualizes metrics such as throughput, latency, and fairness across competing flows. Results show that BBRv3 significantly improves fairness and convergence under AQM, especially with FQ-CoDel. Our visualization tool and modular testbed provide a practical foundation for evaluating next-generation TCP variants in real-world AQM-enabled home wireless networks.

Supervised quantum learning methods face notable limitations in dynamic, real-world environments due to their reliance on static labels and limited adaptability. To address these challenges, we propose a novel online learning framework — Quantum Contextual Bandit (QCB) — that integrates quantum neural networks (QNNs) with contextual bandit (CB) algorithms. The QCB framework enables adaptive decision-making by incorporating bandit-based exploration into QNN training, making it particularly suitable for applications such as recommender systems. To mitigate the adverse effects of quantum noise—including depolarizing, Pauli, and shot noise, the framework leverages a gradient-free optimization approach, enhancing robustness and convergence stability. Experimental results on various datasets demonstrate that QCB consistently outperforms traditional QNN training methods with identical circuit architectures. Notably, the model achieves over 99% accuracy under ideal conditions and sustains high performance under noisy quantum environments. These results underscore the potential of QCB as a scalable, noise-resilient solution for adaptive learning in quantum machine learning systems.

The integration of quantum machine learning into federated learning paradigms is poised to transform the future of technologies that depend on diverse machine learning methodologies. This research delves into Quantum Federated Learning (QFL), presenting an initial framework modeled on the Federated Averaging (FedAvg) algorithm, implemented via Qiskit. Despite its potential, QFL encounters critical challenges, including (i) susceptibility to a single point of failure , (ii) communication bottlenecks, and (iii) uncertainty in model convergence. Subsequently, we dive deeper into QFL and propose an innovative alternative to traditional server-based QFL. Our approach introduces a chained continuous QFL framework (ccQFL), which eliminates the need for a central server and the FedAvg method. In our framework, clients engage in a chained continuous training process, where they exchange models and collaboratively enhance each other’s performance. This approach improves both the efficiency of communication and the accuracy of the training process. Our experimental evaluation includes a proof-of-concept to demonstrate initial feasibility and a prototype study simulating TCP/IP communication between clients. This simulation enables concurrent operations, verifying the potential of ccQFL for real-world applications. We examine various datasets, including Iris, MNIST, synthetic and Genomic, covering a range of data sizes from small to large. For further validity of our proposed method, we extend our experimental analysis in other frameworks such as PennyLane and TensorCircuit where we include various ablation studies covering major considerations and factors that impact the framework to study validity, robustness, practicality, and others. Our results show that the ccQFL framework achieves model convergence, and we evaluate other critical metrics such as performance and communication delay. In addition, we provide a theoretical analysis to establish and discuss many factors such as model convergence, communication costs, etc.

Direct-to-satellite IoT (DtSIoT) enables scalable, global connectivity for diverse applications by leveraging LEO satellites and Non-Orthogonal Multiple Access to support massive device access with heterogeneous QoS needs. However, efficient resource allocation remains a key challenge. To address this, we propose a Deep AutoEncoder-based Model Predictive Controller (DAE-MPC) that learns system dynamics to optimize power allocation in real-time. Simulation results show that DAE-MPC improves transmission efficiency and reduces latency, offering a robust solution for intelligent resource management in DtSIoT networks1.

Through experimental endeavors, we explore the intersection of semantic communication (SemCom) and natural language processing (NLP) to address gaps in SemCom models, focusing on reducing ambiguity and enhancing 6G communication. Our approach involves two phases: 1) Phase 1: Designing an NLP-based system for fair classification, leveraging techniques, such as unsupervised style transfer and zero-shot learning to align human intuition with semantic specifications. 2) Phase 2: Developing modules to minimize and evaluate SemCom performance under channel impairments, integrating language models like DistilBERT and RoBERTa. Results are evaluated using area under the curve receiver operating characteristic (AUC-ROC) metrics across diverse classifiers. Implementation details are publicly available on GitHub. We provide invaluable insights toward learning to harness SemCom with NLP.

Next-generation networks demand security that evolves as fast as threats do. Our pioneering decentralized Zero Trust Architecture (dZTA), proposed in this paper, redefines protection for IoT and remote collaboration, merging Zero Trust’s ironclad access controls with blockchain’s transparency and federated learning’s privacy-first analytics. Unlike traditional models, dZTA enforces security at every layer: a distributed policy engine eliminates single points of failure, cross-network analytics optimize WiFi-8, satellite, and 6G performance under real-world stressors, and anti-leakage protocols safeguard IoT ecosystems. Rigorous real-world simulations confirm dZTA’s dual triumph—uncompromising security and seamless efficiency— proving its readiness to secure tomorrow’s hyperconnected world.

Quantum computing offers unparalleled processing power but raises significant data privacy challenges.Quantum Differential Privacy(QDP) leverages inherent quantum noise to safeguard privacy, surpassing traditional DP. This paper develops comprehensive noise profiles, identifies noise types beneficial for QDP, and highlights the need for practical implementations beyond theoretical models. Existing QDP mechanisms, limited to single noise sources, fail to reflect the multi-source noise reality of quantum systems. We propose a resilient hybrid QDP mechanism utilizing channel and measurement noise, optimizing privacy budgets to balance privacy and utility. Additionally, we introduceLifted Quantum Differential Privacy, offering enhanced randomness for improved privacy audits and quantum algorithm evaluation.

We present a novel decentralized and trustworthy Quantum Federated Learning (QFL) framework tailored for the emerging Metaverse. This virtual environment, enabling social interaction, gaming, and commerce, demands secure and transparent systems. By integrating blockchain, our QFL framework ensures integrity, resilience, and transparency. Comparative analysis with classical Federated Learning (CFL) highlights its practicality and advantages in distributed settings. New insights discovered emphasize the importance of decentralized systems for the Metaverse’s evolution, with a blockchain-based QFL application demonstrated in a hybrid model. Our evaluation, implementation details and code are publicly available.

We develop a novel method by combining weighted personalization with quantum federated averaging to address impending challenges such as non-IID data distribution and client drift. The proposed weighted personalized Quantum Federated Learning (wpQFL) dynamically adapts to data heterogeneity, improving performance, validated through theoretical insights and empirical observations.

A Data-Encoding Approach to Quantum Federated Learning: Experimenting with Cloud Challenges

We investigate the hypothesis that exploring Federated Learning (FL) aggregation methods can enhance training processes within FL frameworks, particularly in resource-constrained edge networks. The methodology employed involved a thorough review of existing FL aggregation methods, leveraging literature databases for data collection and algorithmic simulations for analysis. Distinct taxonomies were introduced to dissect the accuracy and behaviors of these methods. Results revealed critical issues such as communication constraints, personalization, and fairness within FL, emphasizing the necessity for detailed investigations to bridge theory and application gaps. Through meticulous examination and comparative analyses of existing aggregation methods, we provide valuable insights into the development of resilient FL aggregators, laying the groundwork for future advancements in the field.

Improving reliability and response time in decision-making is crucial for efficient demand response (DR) programs in smart grid (SG) environments. By precisely predicting the DR in near real time, consumer premises can be more prepared to optimize energy utilization. We propose and develop an ultrareliable low-latency communication (URLLC)-based machine learning paradigm for orchestrating DR with guaranteed reliability and timeliness. To understand the context in depth and develop new insights, we use a random forest (RF) algorithm to predict the DR program. After that, we employ MuZero reinforcement learning (MuZero RL) on top of RF-based learning with URLLC, which leverages an efficient learned model under such SG DR dynamics. It considerably improves decision-making delays with better generalization to unforeseen situations. In sharp contrast to the state-of-the-art approaches, we observe that MuZero RL enables continuous learning by self-play, achieves higher sample efficiency, and adapts well to the underlying dynamic environments. Such features of an intelligent agent are precious in the context of the considered DR program. We develop theoretical derivations and analyses to study and utilize the framework’s capabilities and demonstrate that URLLC can substantially reduce energy costs.

The Smart Grid (SG) infrastructure plan offers growth opportunities for the electric vehicle (EV) industry and aims to reduce dependence on fossil fuels. Surprisingly, the literature lacks comprehensive research on data privacy issues within the EV-SG ecosystem. In response, this paper presents an efficient federated reinforcement learning (FRL) framework tailored to cost-effectively preserve privacy in wireless vehicle-to-grid (V2G) systems. Our approach involves the use of a small auxiliary battery to generate noise, conceal the true energy demand of electric vehicles, and learn the time-varying dynamics of energy usage for wireless EV charging through a federated process. Within this framework, we employ deep Q-learning to concurrently minimize costs and maximize privacy rewards, while exploring innovative techniques to enhance learning speed and communication efficiency through a global FRL approach.

In Shannon’s theory, semantic aspects of communication were identified but considered irrelevant to the technical communication problems. However, semantic communication techniques have recently attracted renewed research interests in 6G because they have the capability to support an efficient interpretation of the meaning intended by the sender (or accomplishment of the goal) when dealing with multi-dimensional data such as videos, images, audio, sentences, documents etc. We propose a new hashing-based semantic communication, where our learning objective is the "semantic extraction" to produce optimal binary signatures (hash codes) by supervised learning. We evaluate the proposed framework over large image data sets and demonstrate its effectiveness in bulk transmission.

Efficient data communication and machine learning aspects are crucial for orchestrating the distributed resources of smart grid (SG) networks. 5G telecom technologies have enabled ultrareliable low-latency communication (URLLC) to provide low-latency data communication and accelerate distributed machine learning [such as federated learning (FL)] with high reliability. For critical SG operations, such as islanding detection and instability of frequency regulation, the adoption of URLLC and FL appears paramount to enable near real-time communication and collaborative decision making for resource management. However, SG with URLLC and/or FL has been poorly studied in the literature. We develop a novel framework and demonstrate our findings on the importance of URLLC and FL for efficient energy trading between distributed energy sources and to minimize energy loss by enhancing the resilience of critical SG operations. Extensive experiments using real-time data sets validate our design assumptions and ideas.

Future vehicle-to-grid (V2G) systems require more flexible scheduling to adjust and flatten the peak energy demand. For efficient scheduling and energy trading, the utility provider (UP) needs to keep track of the state of charge (SoC) of vehicle batteries (VBs). However, sharing of SoC of VBs from electric vehicles (EVs) to UP may compromise owner privacy by analyzing the electricity usage in EVs. Therefore, we propose Reinforcement learning (RL)-based demand-side energy management using a rechargeable battery (RB) for enhanced cost-friendly privacy of EVs, efficient scheduling, and accurate billing. With existing Q-Learning-based RL (using$\epsilon $-greedy exploration and exploitation), we find that the reward maximization of efficient and private scheduling is often sluggish and incurs convergence issues. Therefore, we develop a genetic algorithm (GA)-based exploration and exploitation, which solves the convergence problems. We develop theoretical analysis and implement numerical results to demonstrate that the proposed GA-based RL framework accelerates convergence and enhances cost-friendly privacy considerably.

Multipath TCP (MPTCP) has emerged as a facilitator for harnessing and pooling available bandwidth in wireless/wireline communication networks and in data centers. Existing implementations of MPTCP such as, Linked Increase Algorithm (LIA), Opportunistic LIA (OLIA) and BAlanced LInked Adaptation (BALIA) include separate algorithms for congestion control and packet scheduling, with pre-selected control parameters. We propose a Deep Q-Learning (DQL) based framework for joint congestion control and packet scheduling for MPTCP. At the heart of the solution is an intelligent agent for interface, learning and actuation, which learns from experience optimal congestion control and scheduling mechanism using DQL techniques with policy gradients. We provide a rigorous stability analysis of system dynamics which provides important practical design insights. In addition, the proposed DQL-MPTCP algorithm utilizes the ‘recurrent neural network’ and integrates it with ‘long short-term memory’ for continuously i) learning dynamic behavior of subflows (paths) and ii) responding promptly to their behavior using prioritized experience replay. With extensive emulations, we show that the proposed DQL-based MPTCP algorithm outperforms MPTCP LIA, OLIA and BALIA algorithms. Moreover, the DQL-MPTCP algorithm is robust to time-varying network characteristics, and provides dynamic exploration and exploitation of paths.

The bottleneck of distributed edge learning (DEL) over wireless has shifted from computing to communication, primarily the aggregation-averaging (Agg-Avg) process of DEL. The existing transmission control protocol (TCP)-based data networking schemes for DEL are application-agnostic and fail to deliver adjustments according to application layer requirements. As a result, they introduce massive excess time and undesired issues such as unfairness and stragglers. Other prior mitigation solutions have significant limitations as they balance data flow rate from workers across paths but often incur imbalanced backlogs when the paths exhibit variance, causing stragglers. To facilitate a more productive DEL, we develop a hybrid multipath TCP (MPTCP) by combining model-based and deep reinforcement learning (DRL) based MPTCP for DEL that strives to realize quicker iteration of DEL and better fairness (by ameliorating stragglers). Hybrid MPTCP essentially integrates two radical TCP developments: i) successful existing model-based MPTCP control strategies and ii) advanced emerging DRL-based techniques, and introduce a novel hybrid MPTCP data transport for easing the communication of Agg-Avg process. Extensive emulation results demonstrate that the proposed hybrid MPTCP can overcome excess time consumption and ameliorate the application layer unfairness of DEL effectively without injecting additional inconstancy and stragglers.

In this article, we consider ultra-reliable low-latency communication (URLLC) for efficient energy trading over a smart grid (SG) network using home-based smart meters (SM). We develop a cost-friendly privacy preservation framework based on existing demand-side energy management by employing random bidirectional energy trading among customers. Customers in our design can be either producers or consumers and mostly both (‘prosumers’). Our aim is to develop a decentralized optimization framework that not only reduces energy costs, but also improves privacy preservation and energy trading ability directly from the customer’s end. One of the vital costs for energy consumers is the supply charge. Our method can minimize it by orchestrating energy trading among customers in a decentralized adaptive fashion. To predict the energy demand by optimizing between privacy and cost, we employ an extension of the follow the regularized leader (FTRL) algorithm. We perform a theoretical analysis to demonstrate the convergence of the FTRL, the benefits of URLLC for the SG network, and the cost-effective privacy preservation ability of the proposed model. In addition to enabling energy trading efficiently, our extensive simulation results demonstrate that our proposed framework outperforms the state-of-the-art methods in terms of the cost-friendly privacy of SMs.

Vanilla Federated learning (FL) relies on the centralized global aggregation mechanism and assumes that all clients are honest. This makes it a challenge for FL to alleviate the single point of failure and dishonest clients. These impending challenges in the design philosophy of FL call for blockchain-based federated learning (BFL) due to the benefits of coupling FL and blockchain (e.g., democracy, incentive, and immutability). However, one problem in vanilla BFL is that its capabilities do not follow adopters’ needs in a dynamic fashion. Besides, vanilla BFL relies on unverifiable clients’ self-reported contributions like data size because checking clients’ raw data is not allowed in FL for privacy concerns. We design and evaluate a novel BFL framework, and resolve the identified challenges in vanilla BFL with greater flexibility and incentive mechanism called FAIR-BFL. In contrast to existing works, FAIR-BFL offers unprecedented flexibility via the modular design, allowing adopters to adjust its capabilities following business demands in a dynamic fashion. Our design accounts for BFL’s ability to quantify each client’s contribution to the global learning process. Such quantification provides a rational metric for distributing the rewards among federated clients and helps discover malicious participants that may poison the global model.

Nowadays, most farming technologies are gradually being transformed into real-time monitoring, control and actuation systems with the advent of the Internet of Things (IoT). These deployments of the IoT over farms have been accelerating due to advancements in sensor technology and communication protocols. Low orbit satellites, for example, has the capability to enable real-time monitoring of the crop over remote places and farms. However, there are numerous challenges to realising seamless farming with LEOs because of low power usage and long-distance transmission requirements from the LoRaIoT sensors over the farms. The main objective of this paper is to study state of the art and present a high-level link budget analysis of the ground sensors and gateways mounted over low earth orbit (LEO) satellites. While the link budget provides ideas on how LEO satellites are prepared for data networking, it also supports improving communication with optimal signal strength. We find that there is a possibility of determining an optimal set of parameters for the ground sensors and the LEO satellite to deliver the desired performance for technical readiness. We observe that different LoRa field parameters such as link budget, receiver power, receiver sensitivity, the path loss can be used at 923.3 MHz frequency. Based on the link budget analysis, we suggest the maximum feasible distance between ground sensors and LEOs. Most importantly, we performed a preliminary analysis of a beamforming approach to improve communication efficiency of the smart farming1.

Enabling ultrareliable low-latency communication (URLLC) with stringent requirements for transmitting data packets (e.g., 99.999% reliability and 1-ms latency) presents considerable uplink transmission challenges. For each packet transmission over dynamically allocated network radio resources, the conventional random access protocols are based on a request-grant scheme. This induces excessive latency and necessitates reliable control signaling, resulting in overhead. To address these problems, grant-free (GF) solutions are proposed in the fifth-generation (5G) new radio (NR). In this article, an overview and vision of the state of the art in enabling GF URLLC are presented. In particular, we first provide a comprehensive review of NR specifications and techniques for URLLC, discuss underlying principles, and highlight impeding issues of enabling GF URLLC. Furthermore, we briefly explain two key phenomena of massive multiple-input–multiple-output (mMIMO) (i.e., channel hardening and favorable propagation) and build several deep insights into how celebrated mMIMO features can be exploited to address the issues and enhance the performance of GF URLLC. Moving further ahead, we examine the potential of cell-free (CF) mMIMO and analyze its distinctive features and benefits over mMIMO to resolve GF URLLC issues. Finally, we identify future research directions and challenges in enabling GF URLLC with CF mMIMO.

We consider future cellular networks and study the coexistence of delay-sensitive (DS) and delay-tolerant (DT) devices in machine-type communication (MTC). DS devices require to minimize access delay for their low delay requirements; in contrast, DT devices have flexible delay constraints. For reducing access delay, we extend fast retrial idea in the data transmissions when a group of preambles is divided into two subsets to support the time-varying nature of the traffic from DS devices. We focus on the stability dynamics by design—we derive convergence conditions by using a control-theoretic approach in terms of variation in the 1) arrival rates; 2) buffer sizing at the devices; 3) number of preambles; and 4) number of DS devices. More importantly, we discover a novel adaptive algorithm that dynamically allocates the number of preambles for DS, thus guaranteeing stability by design. We validate our findings with extensive simulations. Besides, we develop a novel framework that describes how the control theory idea can be applied to address the issue of tracking the buffers and handling coexistence in such a diverse network environment under realistic application constraints. Our extension to the control-theoretic idea predicts whether the overall system is stable, i.e., whether data flows are desynchronized. Given the data flows are desynchronized, small buffers are always sufficient. Our approach shows that a smaller DS buffer often promotes desynchronization—a virtuous cycle.

Abstract In this paper, we perform an in-depth analysis of a large corpus of PDF maldocs to identify the key set of significantly important features and help in maldoc detection. Existing industry-based tools for the detection are inefficient and cannot prevent PDF maldocs because they are generic and depend primarily on a signature-based approach. Besides, several other methods developed by academics suffer heavily from reduced effectiveness. The feature-set using machine learning classifiers is prone to various known attacks, such as mimicry and parser confusion. Also, we discover that increasingly more malicious files i) contain evasive and obfuscated JavaScript code, ii) include hidden contents (mostly outside the objects), iii) have a corrupted document structure, and iv) usually contain short JavaScript code blocks. We utilise maldoc attacks’ evolution over a decade to highlight the essential features (e.g., concept drifts) that impact detectors and classifiers.

Abstract Established methods of communication are based mainly on Shannon’s theory of information, which purposefully overlooks semantic elements of communication. The future wireless technology should promise to facilitate many services, based on content, needs, and semantics, precisely customized to network capabilities. This gave rise to significant concern for Semantic Communication (SC), a novel paradigm considering the message’s meaning during transmission. Federated learning (FL) and Asynchronous Advantage Actor Critic (A3C) are the two emerging distributed and artificially intelligent approaches that provide diverse and possibly massive network coverage for data-driven SC solutions of industry 4.0 automation. Although SC is still in an early development stage, FL-empowered architecture has been recognized as one of the most promising solutions to meet the ubiquitous intelligence in the anticipated sixth-generation (6G) networks. This paper identifies industry 4.0 automation needs that drive the convergence of artificial intelligence and 6G for learning from data streams. We develop a novel SC framework based on the FL and A3C networks and discuss its potential along with transfer learning to address most of the new difficulties anticipated in 6G for industrial communication networks. Our proposed framework has been evaluated with extensive simulation results.

In this article, we develop a novel approach for detecting patients’ position using the radial basis function of the neural network. This new approach aims to continuously monitor the patients’ health statistics and real-time prediction, even when they are outside of cellular coverage. Our research is driven by an initiative to innovate a novel healthcare system of significant importance for intelligent and efficient medical services. For example, doctors need to remotely monitor any patient’s health with the provided health statistics derived from data collected from battery-powered Internet of Things sensors. To this end, our proposed method has been quantified with a holistic mathematical analysis and extensive simulations considering realistic network situations. Our results have accredited the efficiency in the prediction of localization for the patients’ position for the anticipated intelligent healthcare system.

We consider a fifth-generation (5G)-empowered future Industrial IoT (IIoT) networking problem where IIoT machines are capable of communicating and sharing their data networking knowledge gained (and experiences) with other neighboring devices/tools. For such an IIoT setting, deep-learning (DL)-based communication protocols are known to be highly efficient but having a computationally complex training procedure in terms of both time/space and volume of data sets. One solution for such training is to be completed offline for each equipment and machines of IIoT before deployment. A better approach would be to replicate the model from the expert existing machine and implant it into new machines. Such training for the transfer of knowledge can be done by manufacturers using high computational power, even for large-scale DL models. After sufficient training and the desired level of accuracy, the trained machines can be deployed in the smart factory equipment to perform life-long collaborative learning. We design a novel distributed transfer learning (TL) framework to maximize multipath communication networking performance for Industry 4.0 environment. To conduct seamless sharing of knowledge gain by the multipath TCP (MPTCP) agents and tackle retraining issues of DL-based approaches, we investigate TL for MPTCP from the IIoT networking perspective. With relevant insights from transfer and collaborative learning, we develop a distributed TL-MPTCP framework to accelerate the learning efficiency and enhance the performance of newly deployed machines. Our approach is validated with numerical and emulated NS-3 experiments in comparison with the state-of-the-art schemes.

With the increasing adoption of millimeter-waves (mmWave) over cellular networks, outdoor-to-indoor (O2I) communication has been one of the challenging research problems due to high penetration loss of buildings. To address this, we investigate the practicability of utilizing reconfigurable intelligent surfaces (RISs) for assisting such O2I communication. We propose a new notion of prefabricated RIS-empowered wall consisting of a large number of chipless radio frequency identification (RFID) sensors. Each sensor maintains its own bank of delay lines. These sensors which are built within the building walls can potentially be controlled by a main integrated circuit (IC) to regulate the phase of impinging signals. To evaluate our idea, we develop a thorough performance analysis of the RIS-based O2I communication in the mmWave network using stochastic-geometry tools for blockage models. Our analysis facilitates two closed-form approximations of the downlink signal-to-noise ratio (SNR) coverage probability for RIS-based O2I communication. We perform extensive simulations to evaluate the accuracy of the derived expressions, thus providing new observations and findings.

In this article, we design a novel multipath communication framework for Industry 4.0 using deep Q-network [1] to achieve human-level intelligence in networking automation and orchestration. To elaborate, we first investigate the challenges and approaches in exploiting heterogeneous networks and multipath communication [e.g., using multipath transmission control protocol (MPTCP)] for the information technology cum operation technology (IT/OT) convergence in Industry 4.0. Based on the novel idea of intelligent and flexible manufacturing, we analyze the technical challenges of IT/OT convergence and then model network data traffics using MPTCP over the converged frameworks. It quantifies the adverse impact of network convergence on the performance for flexible manufacturing. We provide a few proof-of-concepts solutions; however, after a clear understanding of the tradeoffs, we discover the need for experience-driven MPTCP. The simulation result demonstrates that the proposed scheme significantly outperforms the baseline schemes.

Understanding the performance of compound transmission control protocol (C-TCP) in wireless settings is complicated because of C-TCP's hybrid congestion control, and the complex interdependencies between losses due to wireless channel errors, medium access control (MAC)-layer collisions, and access point (AP) buffer overflows. In this article, we develop a comprehensive model to study the performance of long-lived C-TCP flows over Industry 4.0 WiFi infrastructure, taking all losses into account. Our mathematical model includes WiFi system parameters, such as the retransmissions limit and the AP buffer size, in order to see how they affect transport-layer throughput and fairness. More importantly, we extend the analytical model to multiple APs, and compare the performance of a dual AP scenario with a conventional single AP scenario. Our results show that using cognitive radio and federated learning techniques in the industrial multiple APs scenario can substantially improve the performance.

Massive machine-type Internet of Things (IoT) communication (mMTIC) has the potential for high impact in the anticipated future industry 4.0 sensor networking applications. However, the energy limitation and battery life of the IoT nodes have always been one of the long-standing problems. Clustering routing protocol (CRP) being the most efficient existing approach often suffers when nodes closer to the sink depletes their energy, thereby producing an unwanted energy hole, where packets in flight toward the sink often get interrupted. Considering mMTIC covering a large geographical area, such as monitoring bush fires, the multihop communication among the nodes often causes such an energy hole problem. In this article, we develop an artificial-intelligence-based CRP framework for incorporating a small periphery of a fixed shaped area to ameliorate such energy holes. Our proposed framework is not only energy-optimized but also acts as a robust approach for massive communication and informed data collection.

Cognitive computing, a revolutionary AI concept emulating human brain's reasoning process, is progressively flourishing in the Industry 4.0 automation. With the advancement of various AI and machine learning technologies the evolution toward improved decision making as well as data-driven intelligent manufacturing has already been evident. However, several emerging issues, including the poisoning attacks, performance, and inadequate data resources, etc., have to be resolved. Recent research works studied the problem lightly, which often leads to unreliable performance, inefficiency, and privacy leakage. In this article, we developed a decentralized paradigm for big data-driven cognitive computing (D2C), using federated learning and blockchain jointly. Federated learning can solve the problem of “data island” with privacy protection and efficient processing while blockchain provides incentive mechanism, fully decentralized fashion, and robust against poisoning attacks. Using blockchain-enabled federated learning help quick convergence with advanced verifications and member selections. Extensive evaluation and assessment findings demonstrate D2C's effectiveness relative to existing leading designs and models.

We are introducing an innovative networking approach to the Internet of Vehicles (IoV) for their automation and orchestration by redesigning the wireless edge framework. To this end, we propose to expand the Software Defined Networking (SDN) functionality from the edge network to the cars, i.e. equipping our vehicles with mobile base stations having SDN capabilities. By adopting such a novel approach, most of the impending IoV automation and orchestration challenges are ameliorated, which has been extremely difficult and non-trivial to tackle by using standard existing approaches. The three primary challenges identified in this context are a) the scalability of the network of connected autonomous vehicles, b) the desirable security of the IoV data networking, and c) the flexibility of quality of experience (QoE) perceived by the vehicles. To tackle the aforementioned three challenges, we design a policy-driven framework for a secure and efficient IoV networking paradigm and then investigate its performance via analytic modeling. More importantly, from the IoV data networking perspective, we design an intent-based flow offloading scheme to facilitate enhanced and adjustable QoE. Furthermore, we develop a rigorous analysis to quantify the efficiency of the proposed SDN-capable IoV data networking by modeling TCP connections over WiFi. Using mathematics as a tool for reasoning the three key challenges, our feasibility analysis is validated with extensive simulations. Finally, we provide new insights by deriving the stability conditions for the data flow dynamics of the proposed approach.

Driverless parking, an influential application of Mobility as a Service (MaaS) model, is one of the clear early benefits for autonomous vehicles, given often narrow spaces and multiple potential hazards (such as pedestrians stepping out from in between other vehicles). In recent years, real momentum has been building up for designing automated parking models for vehicles. However, in such an autonomous parking design, location privacy and identity privacy issues are always overlapping due to the improper sharing of data. Most existing studies barely investigate and poorly address such privacy issues. Motivated by this, we develop (and evaluate) an experience-driven, secure and privacy-aware framework of parking reservations for automated cars. Our idea of using differential privacy with zero-knowledge proof provides both security and privacy guarantees to users. Furthermore, the performance of the developed model is enhanced by exploiting reinforcement learning approach such that the utility of the system and the parking reservation rate can be maximized. Extensive evaluation demonstrates the superiority of the proposed model.

The study of optimal properties of centralized dynamic-time-division-duplex (D-TDD) employed at a wireless network consisting of multiple nodes is a highly challenging and partially understood problem in the literature. In this paper, we develop an optimal centralized D-TDD scheme for a wireless network comprised of K full-duplex nodes impaired by self-interference and additive white Gaussian noise. As a special case, we also propose the optimal centralized D-TDD scheme when part or all nodes in the wireless network are half-duplex. Thereby, we derive the optimal adaptive scheduling of the reception, transmission, simultaneous reception and transmission, and silence at every node in the network in each time slot such that the rate region of the network is maximized. The performance of the optimal centralized D-TDD can serve as an upper-bound to any other TDD scheme, which is useful in qualifying the relative performance of TDD schemes. The numerical results show that the proposed centralized D-TDD scheme achieves significant rate gains over existing centralized D-TDD schemes.

With the ensuing surge in data communication volume and the growing need for privacy protection, limiting centralized data collection to the minimum required for specific tasks has been mandatory in industries. This is now guided by the modern privacy legislation, namely, the General Data Protection Regulation and the California Consumer Protection Act. Privacy leakage has become increasingly serious because of massive volume and a variety of data transmission and Quality-of-Service (QoS) requirements in the Industrial Internet-of-Things (IIoT) networks. Although differential privacy is the core privacy protection paradigm, most of its extensions assume all parties share the same level of privacy requirements, which cannot meet varying needs and QoS of IIoT devices in practice. In addition, with multiple paths access to the cloud server (often operated by the trusted third party in IIoT) for higher reliability and performance, satisfying both the privacy and QoS is nontrivial during the data transmission. The usual transmission over both the cellular and WiFi interfaces simultaneously for continuous connectivity among devices, edge networks, and the server is crucial. As a result, we observe that IIoT data privacy is highly vulnerable to collusion attacks. Motivated by this observation, we develop a detailed QoS modeling for multipath TCP over IIoT and propose a QoS-aware personalized privacy protection model. Our model works in two different layers: one at the cloud server and another at the network edges (access points/base station). The aim is not only to balance the load but also to achieve the required QoS and optimize the tradeoff between privacy protection and efficiency. The extensive experimental results based on the real-world data sets illustrate the superiority of the proposed model in terms of privacy protection and efficiency.

Software-defined networks (SDNs) offer unique and attractive solutions to solve challenging management issues in Internet of Things (IoT)-based large-scale multi-technological networks. SDN-IoT network collaboration is innovative and attractive but expected to be extremely heterogeneous in future generation IoT systems. For example, multi-technology network, network externality, and nodes heterogeneity in SDN-IoT may seriously affect the flow or application-specific quality-of-service (QoS) requirements. Furthermore, it highly influences security adoption in a network of interconnected IoT nodes. We observe that both QoS and security are interdependent and nonnegligible factors, thus we emphasize that in order to alleviate heterogeneity it is inevitable to study both these factors hand to hand (or vice versa). With this aim, first, we discuss significant and reasonable cases to encourage researchers to study QoS and security integrally in order to alleviate heterogeneity at SDN-IoT control plane. Second, we propose a framework which successfully transforms the m heterogeneous controllers to n homogeneous controller groups. The key metric of our observation and analysis is the SDN controller's response time. Following this, to validate our approach, we use the mathematical model and a proof of concept (PoC) in a virtual SDN ecosystem is demonstrated. From performance evaluation, we observe that the proposed framework significantly alleviates heterogeneity which helps to maintain QoS and enhance security. This fundamental analysis will enable network security individuals to deal heterogeneity, QoS, and security, of SDN-IoT, in more successful and promising ways.

We propose an autonomous blockchain-based federated learning (BFL) design for privacy-aware and efficient vehicular communication networking, where local on-vehicle machine learning (oVML) model updates are exchanged and verified in a distributed fashion. BFL enables oVML without any centralized training data or coordination by utilizing the consensus mechanism of the blockchain. Relying on a renewal reward approach, we develop a mathematical framework that features the controllable network and BFL parameters (e.g., the retransmission limit, block size, block arrival rate, and the frame sizes) so as to capture their impact on the system-level performance. More importantly, our rigorous analysis of oVML system dynamics quantifies the end-to-end delay with BFL, which provides important insights into deriving optimal block arrival rate by considering communication and consensus delays. We present a variety of numerical and simulation results highlighting various non-trivial findings and insights for adaptive BFL design. In particular, based on analytical results, we minimize the system delay by exploiting the channel dynamics and demonstrate that the proposed idea of tuning the block arrival rate is provably online and capable of driving the system dynamics to the desired operating point. It also identifies the improved dependency on other blockchain parameters for a given set of channel conditions, retransmission limits, and frame sizes.1However, a number of challenges (gaps in knowledge) need to be resolved in order to realise these changes. In particular, we identify key bottleneck challenges requiring further investigations, and provide potential future research directions.1An early version of this work has been accepted for presentation in IEEE WCNC Wksps 2020 [1].

Assuring the required quality of service, despite the growing volume and variety of Internet of Things (IoT) traffic, has remained an immense challenge in the popular WiFi networks. The IoT devices using short TCP flows often attain very different levels of service due to the complicated interactions between the transport layer protocol and the shared dynamic wireless medium. We develop in this paper a novel queue management policy by using a transient model to capture the interactions of IoT (short TCP) flows over traditional traffic in WiFi networks. Based on the Markov regenerative processes coupled with fluid model, we discover that the adaptive admission control (AAC) mechanism at the WiFi access point (AP) improves the response time and fairness of IoT traffic over the lossy WiFi links. To this end, under the proposed AAC, packets are admitted into the AP in such a way that the fractions of packets in the AP buffer belonging to the different IoT devices are balanced resulting in a comparable level of service. We furthermore prove the stability of the system dynamics under AAC which provides important practical insights in designing home WiFi IoT system.

We propose to redesign SDN control in IoT lastmiles so as to extend the capability from edge routers to devices (end-node things enabled with SDN capabilities). Our approach put forward existing and new challenges that are impossible to be resolved using the seminal approaches directly. The main challenges we identify are: scalability of sensor nodes/things, maintaining the security of the system, and fulfilling the Quality of Service (QoS) requirement of all IoT applications. Firstly, we elaborate and discuss the aforementioned critical and fundamental challenges that require immediate investigations. Secondly, we propose a policy-driven framework for secure routing and conduct performance modeling and analysis. Further, in the QoS context, we have proposed an intent-based flow offloading scheme to meet the flow-specific QoS requirements. More importantly, we have developed an analysis by modeling TCP-based flows over WiFi, thus forming the required SDN-IoT network, by using mathematics as a tool for reasoning our challenges. With new insights from our analysis, the feasibility of the proposed approach is validated using factors such as path set-up time in SDN-IoT networks, SDN controller/devices throughputs, packets losses and response time of the controller.

The proliferation of devices with multiple wireless interfaces and (automobile) manufacturers' interest in implementing reliable connectivity for vehicles create an ideal scenario for multipath transmission control protocol (MPTCP) over vehicular networks. Next generation vehicular networks will be highly dynamic and exploit MPTCP over various wireless technologies including WiFi and 5th generation (5G) cellular networks to form the Internet of vehicles (IoV). However, the probability of packet loss can be high and/or packets arrive at the destination out-of-order due to time-varying heterogeneous wireless paths. Further, the IoV traffic is delay sensitive, which urges the need to investigate MPTCP algorithms for reliable communication over heterogeneous lossy networks, while satisfying delay constraints. We approach these challenges by jointly using load balancing and forward error correction (FEC) for performing coupled congestion control inside MPTCP. As a result, the proposed MPTCP-IoV is TCP-friendly by design, and provably stable and convergent to a unique equilibrium point. In addition, a comprehensive mathematical analysis (developed to approximate the reordering delay) is carried out and shown that the delay penalty for MPTCP-IoV over such networks becomes insignificant when considering the potential gains obtained by the convergence of multiple networks. Experiments are carried out and the results demonstrate these characteristics.

A complete understanding of the dynamics of Multipath TCP (MPTCP) over Cellular and WiFi networks is still lacking. This is a highly challenging issue, as a consequence of the complex interdependencies between the losses, packet reordering due to heterogeneous wireless channel features, errors, and link layer retransmissions, as well as their (joint) influence on MPTCP's control mechanism. In this paper, we develop a comprehensive approach that is capable of assessing the performance of long-lived MPTCP flows with joint WiFi and Cellular network access, taking into account the diverse characteristics of both types of networks. Relying on a parallel queueing model, we develop a framework that features the controllable network parameters, such as the retransmission limit and the buffer sizes, so as to capture their impact on the TCP-level performance. We include a variety of numerical and simulation results highlighting various non-trivial findings and insights for adaptive MPTCP design. In particular, we design a novel MPTCP algorithm to exploit the route heterogeneity, and demonstrate that the proposed algorithm is capable of learning the network's characteristics. It also identifies the improved MPTCP window increment parameters for any given set of channel errors, retransmission limits, and buffer sizes.

Recent advancements in Internet congestion control have introduced a multipath TCP (MPTCP) that aims to simultaneously utilize multiple available paths in the network. In this paper, we develop an integrated fluid and packet-level analytical model to study the coexistence of regular and MPTCP users sharing a common WiFi access point (AP). We observe a throughput unfairness of MPTCP with regular TCP in the last-mile WiFi networks. In order to fix the fairness issue, we develop a real-time Adaptive Loss Management (ALM) algorithm that continuously monitors the deviation in AP buffer occupancy and adapts its packet admission probability based on a closed form expression derived from our analytical model. We provide a proof as well as show via numerical and simulation results that the proposed ALM algorithm is TCP-friendly by design, and provably stable.

We develop a comprehensive analytical model for multiple long-lived multipath Transmission Control Protocol (TCP) connections downloading content from a remote server in the Internet using parallel paths with Wi-Fi and cellular last-miles. This is the first analytical model developed in the literature that captures the coupling between the paths through heterogeneous wireless networks where the coupling arises due to the multipath TCP coupled congestion control protocol. The model also takes into account the impact of the shared nature of the wireless medium and the finite access point (AP) buffer in the Wi-Fi last-mile. The accuracy of the proposed model is demonstrated via extensive ns-2 simulations. Furthermore, we discover a new type of throughput unfairness among the competing regular and multipath TCP connections going through the same AP with a droptail buffer; the regular TCP connections essentially steal almost all the Wi-Fi bandwidth away from the multipath TCP connections. To tackle this problem, we present two simple solutions utilizing our analytical model and achieve fairness.

We propose an analytical model for a Wi-Fi network acting as a last-mile Internet access with multiple long-lived TCP connections on both the up and down links. Our model considers the joint impact of buffer losses at the access point, contention at the medium access control layer, and packet losses due to the wireless channel being erroneous. We show that the model accurately quantifies the probability of an arbitrary TCP packet being discarded, and the total throughput obtained on the up and down links. Furthermore, quantitative insights can be gained into the throughput that long-lived TCP flows achieve under the joint impact of all aforementioned types of losses. In particular, we find that the wireless channel errors and buffer overflows both lead to throughput unfairness, but that they do so in the opposite direction on the up and down links, respectively. We demonstrate that this insight can be exploited so as to significantly mitigate the throughput unfairness without compromising the total obtainable network throughput.

TCP NewReno is one of the most widely deployed TCP versions in today's Internet. However, a full understanding of the complex inter-dependencies between the losses due to wireless channel errors and those due to buffer overflows, and their (joint) impact on TCP NewReno's congestion control algorithm in wireless and wired-cum-wireless networks is still lacking. In this paper, we develop a comprehensive analytical model for, and study the performance of, TCP NewReno with a cellular last-mile access, taking into account both types of losses. We assume a frame-level Markovian loss model, and build a model that features the system's basic controllable parameters (such as the number of retransmissions and the buffer size), so as to study how they (jointly) affect the TCP-level throughput. We model certain finer aspects, e.g., correlations in wireless and buffer losses and their cross-correlation. We provide a summary of numerical results highlighting several non-trivial findings. In particular, we demonstrate that there exist optimal (i.e., TCP throughput maximizing) pairs of the number of retransmissions and the buffer size.